mRNA 3'UTR调控肿瘤竞争性内源RNA网络的研究

A recent breakthrough in microRNA mediated gene regulation is the establishment of "competing endogenous RNA" (ceRNA) mechanism. Through ceRNA regulatory network, RNAs sharing binding sites for the same microRNA regulate each other's expression by competing for a limited pool of microRNAs.Since the microRNA binding sites are mainly located in the 3'UTR region of mRNAs and it has been reported that oncogenes in cancer cell lines typically express shorter UTRs, we hypothesized that the alternative splicing of mRNA 3'UTR could have a significant impact on the ceRNA regulatory network. In earlier works, we have established an Bayesian based approach to calculate the changes of mRNA 3'UTR length using Affymetrix human Exon array data. Our analysis of a cohort of 185 prostate cancer exon arrays showed that not only there are systematic level change of mRNA 3'UTR lengths, but those changes also have excellent performance in identifying high risk patients for cancer recurrence.In a follow up analysis, we also showed that 3'UTR length dynamics has a significant impact on known ceRNA regulatory networks. In this proposed study, we will further carry out a large scale meta-analysis using the TCGA ovarian cancer, the TCGA Glioblastoma multiforme and the MSKCC prostate cancer data to systematically analyze the effects of 3'UTR length dynamics on ceRNA regulatory networks. The results of our analysis could reveal insightful mechanisms of gene regulation by connecting alternative splicing with ceRNA regulatory networks. The mechanism could also shed lights into the tumorigenesis mechanism and help to establish novel algorithms for tumor classification and prognosis prediction.

最近的肿瘤系统生物学研究提出了"竞争性内源RNA"的理论（ceRNA），并确认了多个癌基因的mRNA能够通过竞争miRNA来互相调控。基于肿瘤细胞中mRNA 3' UTR由于选择性剪接而发生长短变化的现象，本课题提出了肿瘤特异的3' UTR长短变化会引发ceRNA调控网络变化，进而影响肿瘤发生发展的假说。在前期工作中，我们建立了通过Exon array数据来决定3' UTR长短变化的新算法。对前列腺癌的数据分析证实了3' UTR长短变化能高效预测恶性肿瘤复发。 进一步的分析初步证实了肿瘤3'UTR长短变化对调控重要癌基因的ceRNA网络有显著影响。在此基础上，我们会整合多种肿瘤基因组数据，对本假说进行计算分析及生物验证，并建立新的肿瘤预测模型。本课题提出和验证并将继续阐明一种新的基因调控机制。本项目成果既能加深对肿瘤基因调控的理解，又将有助于创建新的肿瘤个性化治疗预测工具。

竞争内源性RNA（ceRNA）已经被确认为调节基本生物过程的通用机制。许多研究表明，ceRNA的失调是肿瘤发生的关键机制。然而，现有的研究主要集中在单个ceRNA分子，并且很少知道在肿瘤发生期间的系统性失调的ceRNA网络和这种系统性失调在预后中的效用。在这项研究中，我们开发了一种新的基于贝叶斯断点的方法，从微阵列的探针水平数据估计3'UTR缩短，系统分析了3'UTR缩短对ceRNA网络的影响及其在多种癌症中的预后效用。我们的研究结果表明3'UTR缩短是肿瘤发生期间普遍观察到的现象，并可以从微阵列的探针水平数据有效地推断。关键的是，3'UTR缩短诱导ceRNAs网络在肿瘤发生期间发生重大失调。并且失调的严重性与癌症的进展密切相关，其中最严重的变化发生在晚期和转移性肿瘤中。关键的是，ceRNA网络上的失调高度特异性的针对关键癌基因，并在很大程度上造成相应的异常基因表达。最后，3'UTR缩短谱在将肿瘤样品分成具有不同临床结果的组中是高度有效的。这些结果表明ceRNA网络的特异性失调是肿瘤发生的基本特征，并建立了3'UTR缩短作为将肿瘤患者分类为临床有意义亚型的有效方法。